Researchers from LMU Munich and the University of Baghdad have employed artificial intelligence to reconstruct a poem titled “Babylonian Hymn,” lost for over 2,000 years, reports Decrypt.
The text praises Babylon and the god Marduk. It was composed 3,000 years ago and last studied in 100 BC.
The team behind the project revealed that the work was assembled from 30 clay fragments unearthed over many years, with AI used to piece the work together.
“We used a specialized AI-based program to analyze and match text fragments by combinations of cuneiform signs,” said Enrique Jiménez, professor of Ancient Near Eastern Languages at LMU Munich.
The expert, along with colleagues, employs methods based on natural language processing to determine which text the fragments belong to.
Working with the Electronic Babylonian Library platform, which contains 1,402 manuscripts, researchers use n-gram matching as the primary method of reconstruction. Other approaches include overlapping vocabulary and searching for the longest common sequences of text.
Jiménez emphasized that the discovered poem was significant for its contemporaries and was part of Babylon’s educational curriculum.
In an article for the journal Iraq, he and co-author Anmar Fadhil suggested that the poem’s creator likely belonged to Babylon’s priestly class, as one section describes priests as “free citizens.”
The hymn extols the city’s natural resources and beauty, respect for foreigners, and support for the poor.
“They do not humiliate foreigners among them. They protect the humble, support the weak. Under their care, the poor and destitute can thrive. They grant orphans comfort and favor,” the composition states.
Jiménez noted that artificial intelligence is becoming indispensable for researchers, “especially for restoring damaged or fragmented texts.”
“Although languages like Akkadian and Sumerian are still underrepresented in LLM, we are actively working on improving computational tools for the study of the ancient Near East,” the expert noted.
Previously, scientists from DeepMind and Ca’ Foscari University of Venice developed the AI algorithm Ithaca for restoring damaged ancient texts.
Back in October 2021, specialists from Google AI recreated destroyed paintings by Gustav Klimt using artificial intelligence.